Tootfinder

NFL Draft prospect Rueben Bain Jr. involved in fatal car accident in 2024, per report

https://www.cbssports.com/nfl/news/nfl-draft-prospect-rueben-bain-jr-fatal-car-accident/

just dislodged a memory of attending the original 2002 auction of jerry garcia's guitars, at studio 54 for some reason, & guitar maker doug irwin (who garcia willed them to) literally cackling/laughing from the back of the room when tiger went for almost a million.

So to follow up on this, I've caught it in action. Models, when quantized a bit, just do a bit more poorly with short contexts. Even going from f32 (as trained) to bf16 (as usually run) to q8 tends to do okay for "normal" context windows. And q4 you start feeling like "this model is a little stupid and gets stuck sometimes” (it is! It's just that it's still mostly careening about in the space of "plausible" most of the time. Not good guesswork, but still in the zone). With long contexts, the probability of parameters collapsing to zero are higher, so the more context the more likelihood you are to see brokenness.
And then at Q2 (2 bits per parameter) or Q1, the model falls apart completely. Parameters collapse to zero easily. You start seeing "all work and no play makes jack a dull boy” sorts of behavior, with intense and unscrutinized repetition, followed by a hard stop when it just stops working.
And quantization is a parameter that a model vendor can turn relatively easily. (they have to regenerate the model from the base with more quantization, but it's a data transformation on the order of running a terabyte through a straightforward and fast process, not like training).
If you have 1000 customers and enough equipment to handle the requests of 700, going from bf16 to q8 is a no-brainer. Suddenly you can handle the load and have a little spare capacity. They get worse results, probably pay the same per token (or they're on a subscription that hides the cost anyway so you are even freer to make trade-offs. There's a reason that subscription products are kinda poorly described.)
It's also possible for them to vary this across a day: use models during quieter periods? Maybe you get an instance running a bf16 quantization. If you use it during a high use period? You get a Q4 model.
Or intelligent routing is possible. No idea if anyone is doing this, but if they monitor what you send a bit, and you generally shoot for an expensive model for simple requests? They could totally substitute a highly quantized version of the model to answer the question.
There are •so many tricks• that can be pulled here. Some of them very reasonable to make, some of them treading into outright misleading or fraudulent, and it's weirdly hard to draw the line between them.

Raiders’ Ashton Jeanty Shares True Feelings on Mike Washington Jr. During Minicamp https://heavy.com/sports/nfl/las-vegas-raiders/ashton-jeanty-mike-washington-feelings-minicamp/

"Ideally, that's what you've got in an acting career is an equal number of dramas and comedies and an equal number of small films and big films."
—Jason Bateman
#acting #coaching #inspiration

Since falling in love with soccer during Euro 2000, I've never been so physically close to the World Cup (the first match in Toronto is later today). But I've never felt so distant from the competition. Few organizations are as good at finding new ways to disgust as FIFA and the US regime's actions - not just those focused on the World Cup, but those are bad enough - fill me with rage. In the past, I've plugged my nose and enjoyed the matches (I'm not proud). I can't …

"Toby Carvery to fund orchard replanting as settlement for felling ancient oak"
#UK #UnitedKingdom #Trees #Environment

Das #Landgericht Flensburg stuft die #LetzteGeneration nicht als #kriminelleVereinigung ein.
Straßenblockaden und andere Protestformen erfüllten nicht die Voraussetzung…

Urteil des Landgerichts Flensburg: Letzte Generation keine kriminelle Vereinigung
Das Landgericht Flensburg sieht in der Letzten Generation keine kriminelle Vereinigung. Die Aktionen seien keine Gefahr für die öffentliche Sicherheit.

"Perseverance is failing 19 times and succeeding the 20th."
—Julie Andrews
#acting #coaching #inspiration

"Ideally, that's what you've got in an acting career is an equal number of dramas and comedies and an equal number of small films and big films."
—Jason Bateman
#acting #coaching #inspiration